def test_amber_binary_mixture():
sustiva_filename = utils.get_data_filename("chemicals/etoh/etoh.mol2")
etoh_filename = utils.get_data_filename("chemicals/etoh/etoh_renamed.mol2")
trj0, trj1 = md.load(sustiva_filename), md.load(etoh_filename)
# Hack to assign unique residue names that are consistent with contents of mol2 files
trj0.top.residue(0).name = "LIG"
trj1.top.residue(0).name = "LI2"
trj_list = [trj0, trj1]
with utils.enter_temp_directory(): # Prevents creating tons of GAFF files everywhere.
box_filename = "./box.pdb"
box_trj = packmol.pack_box(trj_list, [25, 50])
box_trj.save(box_filename)
gaff_mol2_filename0, frcmod_filename0 = utils.run_antechamber("sustiva", sustiva_filename, charge_method=None)
gaff_mol2_filename1, frcmod_filename1 = utils.run_antechamber("etoh", etoh_filename, charge_method=None)
mol2_filenames = [gaff_mol2_filename0, gaff_mol2_filename1]
frcmod_filenames = [frcmod_filename0, frcmod_filename1]
prmtop_filename = "./out.prmtop"
inpcrd_filename = "./out.inpcrd"
tleap_cmd = amber.build_mixture_prmtop(mol2_filenames, frcmod_filenames, box_filename, prmtop_filename, inpcrd_filename)
print(tleap_cmd)
def test_run_antechamber_charges():
molecule_name = "acetate"
input_filename = utils.get_data_filename("chemicals/acetate/acetate.mol2")
with utils.enter_temp_directory(): # Prevents creating tons of GAFF files everywhere.
gaff_mol2_filename, frcmod_filename = utils.run_antechamber(
molecule_name, input_filename, charge_method=None, net_charge=-1
)
def test_acpype_conversion():
molecule_name = 'sustiva'
input_filename = utils.get_data_filename("chemicals/sustiva/sustiva.mol2")
with utils.enter_temp_directory(): # Prevents creating tons of GAFF files everywhere.
gaff_mol2_filename, frcmod_filename = utils.run_antechamber(molecule_name, input_filename, charge_method=None)
prmtop, inpcrd = utils.run_tleap(molecule_name, gaff_mol2_filename, frcmod_filename)
out_top, out_gro = utils.convert_via_acpype( molecule_name, prmtop, inpcrd )
def oemols_to_ffxml(molecules, base_molecule_name="lig"):
"""Generate an OpenMM ffxml object and MDTraj trajectories from multiple OEMols
Parameters
----------
molecules : list(OEMole)
Molecules WITH CHARGES. Each can have multiple conformations.
WILL GIVE UNDEFINED RESULTS IF NOT CHARGED.
base_molecule_name : str, optional, default='lig'
Base name of molecule to use inside parameter files.
Returns
-------
trajectories : list(mdtraj.Trajectory)
List of MDTraj Trajectories for molecule. May contain multiple frames
ffxml : StringIO
StringIO representation of ffxml file.
Notes
-----
We allow multiple different molecules at once so that they can all be
included in a single ffxml file, which is currently the only recommended
way to simulate multiple GAFF molecules in a single simulation. For most
applications, you will have just a single molecule:
e.g. molecules = [my_oemol]
The resulting ffxml StringIO object can be directly input to OpenMM e.g.
`forcefield = app.ForceField(ffxml)`
This will generate a lot of temporary files, so you may want to use
utils.enter_temp_directory() to avoid clutter.
"""
all_trajectories = []
gaff_mol2_filenames = []
frcmod_filenames = []
print(os.getcwd())
for i, molecule in enumerate(molecules):
trajectories = []
for j in range(molecule.NumConfs()):
molecule_name = "%s-%d-%d" % (base_molecule_name, i, j)
mol2_filename = "./%s.mol2" % molecule_name
_unused = molecule_to_mol2(molecule, mol2_filename, conformer=j)
gaff_mol2_filename, frcmod_filename = run_antechamber(
molecule_name, mol2_filename, charge_method=None
) # It's redundant to run antechamber on each frame, fix me later.
traj = md.load(gaff_mol2_filename)
trajectories.append(traj)
if j == 0: # Only need 1 frame of forcefield files
gaff_mol2_filenames.append(gaff_mol2_filename)
frcmod_filenames.append(frcmod_filename)
# Create a trajectory with all frames of the current molecule
traj = trajectories[0].join(trajectories[1:])
all_trajectories.append(traj)
ffxml = create_ffxml_file(gaff_mol2_filenames, frcmod_filenames, override_mol2_residue_name=base_molecule_name)
return all_trajectories, ffxml
def smiles_to_antechamber(smiles_string,
gaff_mol2_filename,
frcmod_filename,
residue_name="MOL",
strictStereo=False):
"""Build a molecule from a smiles string and run antechamber,
generating GAFF mol2 and frcmod files from a smiles string. Charges
will be generated using the OpenEye QuacPac AM1-BCC implementation.
Parameters
----------
smiles_string : str
Smiles string of molecule to construct and charge
gaff_mol2_filename : str
Filename of mol2 file output of antechamber, with charges
created from openeye
frcmod_filename : str
Filename of frcmod file output of antechamber. Most likely
this file will be almost empty, at least for typical molecules.
residue_name : str, optional, default="MOL"
OpenEye writes mol2 files with <0> as the residue / ligand name.
This chokes many mol2 parsers, so we replace it with a string of
your choosing. This might be useful for downstream applications
if the residue names are required to be unique.
strictStereo : bool, optional, default=False
If False, permits smiles strings with unspecified stereochemistry.
See https://docs.eyesopen.com/omega/usage.html
"""
oechem = import_("openeye.oechem")
if not oechem.OEChemIsLicensed():
raise (ImportError("Need License for oechem!"))
# Get the absolute path so we can find these filenames from inside a temporary directory.
gaff_mol2_filename = os.path.abspath(gaff_mol2_filename)
frcmod_filename = os.path.abspath(frcmod_filename)
m = smiles_to_oemol(smiles_string)
m = get_charges(m, strictStereo=strictStereo, keep_confs=1)
with enter_temp_directory(
): # Avoid dumping 50 antechamber files in local directory.
_unused = molecule_to_mol2(m, "./tmp.mol2", residue_name=residue_name)
net_charge = oechem.OENetCharge(m)
tmp_gaff_mol2_filename, tmp_frcmod_filename = run_antechamber(
"tmp", "./tmp.mol2", charge_method=None,
net_charge=net_charge) # USE OE AM1BCC charges!
shutil.copy(tmp_gaff_mol2_filename, gaff_mol2_filename)
shutil.copy(tmp_frcmod_filename, frcmod_filename)
def smiles_to_antechamber(smiles_string, gaff_mol2_filename, frcmod_filename, residue_name="MOL", strictStereo=False):
"""Build a molecule from a smiles string and run antechamber,
generating GAFF mol2 and frcmod files from a smiles string. Charges
will be generated using the OpenEye QuacPac AM1-BCC implementation.
Parameters
----------
smiles_string : str
Smiles string of molecule to construct and charge
gaff_mol2_filename : str
Filename of mol2 file output of antechamber, with charges
created from openeye
frcmod_filename : str
Filename of frcmod file output of antechamber. Most likely
this file will be almost empty, at least for typical molecules.
residue_name : str, optional, default="MOL"
OpenEye writes mol2 files with <0> as the residue / ligand name.
This chokes many mol2 parsers, so we replace it with a string of
your choosing. This might be useful for downstream applications
if the residue names are required to be unique.
strictStereo : bool, optional, default=False
If False, permits smiles strings with unspecified stereochemistry.
See https://docs.eyesopen.com/omega/usage.html
"""
oechem = import_("openeye.oechem")
if not oechem.OEChemIsLicensed():
raise (ImportError("Need License for oechem!"))
# Get the absolute path so we can find these filenames from inside a temporary directory.
gaff_mol2_filename = os.path.abspath(gaff_mol2_filename)
frcmod_filename = os.path.abspath(frcmod_filename)
m = smiles_to_oemol(smiles_string)
m = get_charges(m, strictStereo=strictStereo, keep_confs=1)
with enter_temp_directory(): # Avoid dumping 50 antechamber files in local directory.
_unused = molecule_to_mol2(m, "./tmp.mol2", residue_name=residue_name)
net_charge = oechem.OENetCharge(m)
tmp_gaff_mol2_filename, tmp_frcmod_filename = run_antechamber(
"tmp", "./tmp.mol2", charge_method=None, net_charge=net_charge
) # USE OE AM1BCC charges!
shutil.copy(tmp_gaff_mol2_filename, gaff_mol2_filename)
shutil.copy(tmp_frcmod_filename, frcmod_filename)
def test_amber_box():
etoh_filename = utils.get_data_filename("chemicals/etoh/etoh.mol2")
trj_list = [md.load(etoh_filename)]
with utils.enter_temp_directory(): # Prevents creating tons of GAFF files everywhere.
box_filename = "./box.pdb"
box_trj = packmol.pack_box(trj_list, [50])
box_trj.save(box_filename)
gaff_mol2_filename1, frcmod_filename1 = utils.run_antechamber("etoh", etoh_filename, charge_method=None)
mol2_filenames = [gaff_mol2_filename1]
frcmod_filenames = [frcmod_filename1]
prmtop_filename = "./out.prmtop"
inpcrd_filename = "./out.inpcrd"
tleap_cmd = amber.build_mixture_prmtop(mol2_filenames, frcmod_filenames, box_filename, prmtop_filename, inpcrd_filename)
print(tleap_cmd)
def test_run_tleap():
molecule_name = "sustiva"
input_filename = utils.get_data_filename("chemicals/sustiva/sustiva.mol2")
with utils.enter_temp_directory(): # Prevents creating tons of GAFF files everywhere.
gaff_mol2_filename, frcmod_filename = utils.run_antechamber(molecule_name, input_filename, charge_method=None)
prmtop, inpcrd = utils.run_tleap(molecule_name, gaff_mol2_filename, frcmod_filename)
def oemols_to_ffxml(molecules, base_molecule_name="lig"):
"""Generate an OpenMM ffxml object and MDTraj trajectories from multiple OEMols
Parameters
----------
molecules : list(OEMole)
Molecules WITH CHARGES. Each can have multiple conformations.
WILL GIVE UNDEFINED RESULTS IF NOT CHARGED.
base_molecule_name : str, optional, default='lig'
Base name of molecule to use inside parameter files.
Returns
-------
trajectories : list(mdtraj.Trajectory)
List of MDTraj Trajectories for molecule. May contain multiple frames
ffxml : StringIO
StringIO representation of ffxml file.
Notes
-----
We allow multiple different molecules at once so that they can all be
included in a single ffxml file, which is currently the only recommended
way to simulate multiple GAFF molecules in a single simulation. For most
applications, you will have just a single molecule:
e.g. molecules = [my_oemol]
The resulting ffxml StringIO object can be directly input to OpenMM e.g.
`forcefield = app.ForceField(ffxml)`
This will generate a lot of temporary files, so you may want to use
utils.enter_temp_directory() to avoid clutter.
"""
all_trajectories = []
gaff_mol2_filenames = []
frcmod_filenames = []
print(os.getcwd())
for i, molecule in enumerate(molecules):
trajectories = []
for j in range(molecule.NumConfs()):
molecule_name = "%s-%d-%d" % (base_molecule_name, i, j)
mol2_filename = "./%s.mol2" % molecule_name
_unused = molecule_to_mol2(molecule, mol2_filename, conformer=j)
gaff_mol2_filename, frcmod_filename = run_antechamber(
molecule_name, mol2_filename, charge_method=None
) # It's redundant to run antechamber on each frame, fix me later.
traj = md.load(gaff_mol2_filename)
trajectories.append(traj)
if j == 0: # Only need 1 frame of forcefield files
gaff_mol2_filenames.append(gaff_mol2_filename)
frcmod_filenames.append(frcmod_filename)
# Create a trajectory with all frames of the current molecule
traj = trajectories[0].join(trajectories[1:])
all_trajectories.append(traj)
ffxml = create_ffxml_file(gaff_mol2_filenames,
frcmod_filenames,
override_mol2_residue_name=base_molecule_name)
return all_trajectories, ffxml
#!/usr/bin/env python
import sys
from openmoltools.utils import run_antechamber
if __name__ == "__main__":
if len(sys.argv) <= 1:
print("""Usage: generate_example_data.py ligand_name
Note: this should be run in the openmoltools/chemicals/ligand_name directory.
""")
else:
molecule_name, mol2_filename = sys.argv[1:]
run_antechamber(molecule_name, mol2_filename, charge_method="bcc")
